Private infrastructure for cloud natives.

amd-gpu-operator - AMD GPU supportcert-manager - certificate managementlonghorn - replicated storageminio - S3-compatible storage (used as Longhorn backup target)nfd - Node Feature Discovery (GPU autodetection)nvidia-gpu-operator - NVidia GPU supporttailscale-operator - ingress support with Tailscale authenticationtraefik - reverse proxy for custom domain supportcloudnative-pg - PostgreSQL operatormariadb-operator - MariaDB operatorbeszel - Beszel lightweight monitoringprometheus - Prometheus/Grafana monitoringhome-assistant - sensor and home automation platformautomatic1111 - Automatic1111 Stable Diffusion WebUIkubeai - Ollama and vLLM models over OpenAI-compatible APIinvokeai - generative AI plaform, community editionn8n - AI workflow automationollama - local large language modelsopen-webui - Open WebUI frontendsdnext - SD.Next Stable Diffusion WebUIbitcoin-core - Bitcoin Core nodebitcoin-knots - Bitcoin Knots nodeelectrs - Electrs (Electrum) server implementationmempool - Blockchain explorernextcloud - File sharing and collaboration suiteInstallation instructions assume your machines are running Bluefin (Developer edition, https://projectbluefin.io/) based on Fedora Silverblue unless otherwise noted. It should run on any modern Linux distribution with Linux kernel 6.11.6+, even including Raspberry Pi.
Windows and MacOS support is limited, specifically they cannot be used as storage nodes.
See Disabling Longhorn Guide with instructions on using local-path-provisioner instead of Longhorn.
Both NVIDIA and AMD GPUs are supported. See AMD GPU support for more information.
After system components have been deployed, you can add any of the optional #Applications. Details in each module documentation.
For general application configuration and deployment instructions, see Configuration Guide.
$ claude mcp add orange-lab \
-- python -m otcore.mcp_server <graph>